AI Breakout

A Deep Q-Learning (DQN) implementation for the classic Atari Breakout game, using RAM state representation and advanced visualization tools.

Overview

This project implements a Deep Q-Network (DQN) agent that learns to play the Atari Breakout game using RAM state representation. The implementation includes comprehensive visualization and analysis tools to understand the agent's decision-making process. For the full project details, reference the Final Paper

Requirements

• Python 3.8+
• PyTorch
• Gymnasium
• ALE-py
• Matplotlib
• NumPy

Installation

1. Clone the repository:

git clone https://github.com/yourusername/ai-breakout.git
cd ai-breakout

2. Create and activate a virtual environment (recommended):

python -m venv env
source env/bin/activate  # On Windows: env\Scripts\activate

3. Install dependencies:

pip install -r requirements.txt

4. Install Atari ROMs:

python -m ale_py.roms.install

Project Structure

• visualize_agent.py: Main visualization and analysis script
• ram_state_representation/: Contains the DQN agent implementation
• evaluate.py: Evaluation script for agent performance
• test_environment.py: Environment testing utilities